Privatesector P U B L I C P O L I C Y F O R T H E The World Bank FPD Note No. 10 June 1994 Restructuring the Power Sector: The Case of Small Systems Robert Bacon New paradigm system. But that is not the advice The recent worldwide interest in pow- some countries are getting. For exam- er sector privatization and restructur- ple, a consultant report for Kenya, ing has focused on a few high-profile which had net installed public capacity cases, such as Argentina, Chile, and of 706 megawatts (MW) in 1990, said the United Kingdom. As a result, the that there was no scope for competi- pattern of restructuring now taking tion between the existing generating shape in many power systems largely plants. Yet consultants have still reflects the experience in these few advised separating the single private countries. Consultants working in loca- generator from the transmission and tions as diverse as Jamaica, Kenya, distribution company. and Poland are applying a privatiza- tion model in which generation is sep- The arguments for this vertical separa- arated from transmission, which in tion are quite different from those for turn may be separated from distribu- horizontal separation and need to be tion. The generation sector is then individually addressed. Indeed, even split into several competing firms. This when it is possible to introduce limited model of restructuring, based on cap- competition in generation and hence turing the benefits of competition in achieve some benefits, the costs of the generation sector, is accompanied vertical separation may be so large as by regulation for those parts of the to offset the gains from competition. new system that cannot be competitive Therefore, any restructuring plan for a and that may therefore open up possi- small system must take into account bilities of monopoly exploitation. that it may be harder to achieve real competition in the generation sector. Relevance to small systems In small power systems, however, the This issue is relevant to many small balance of advantages and disadvan- countries contemplating power sector tages from these changes may be quite reform. In 1990, there were sixty coun- different from that of systems in larger tries with capacity of less than 150 economies. When there is only a sin- MW, thirty with a total net public ca- gle generator, it can be more efficient pacity of between 150 and 500 MW, to leave it joined to the transmission and seventeen with between 500 and Industry and Energy Department Vice Presidency for Finance and Private Sector Development 1,000 MW. The experience of countries that have al- ship with the company. Where this is more difficult, ready restructured their power sectors (United King- because of the political situation or because of the dom, system size 70,000 MW; Argentina, 15,000 MW; traditional approach to state companies, privatization and Chile, 3,000 MW) may be of very limited relevance may bring permanent benefits that would not be sus- to systems of, say, 1,000 MW or less. tainable with a commercialized state entity. The most efficient structure for a small system may The allocative gain or loss from a change in pricing be quite different from the fully disaggregated mod- due to a shift from a state monopoly to a private el. Reform proposals thus need to be flexible, and monopoly depends on the extent to which prices alternative systems under consideration need to be were being held below costs in the first case and the closely evaluated. This Note reviews the sources of extent to which prices are limited by regulation in the kinds of gains and losses from privatization and the second. With no subsidization, a move to unreg- restructuring that need to be considered in any pro- ulated pricing by a private monopoly produces a posed power sector reform strategy. It also looks at “deadweight” loss of consumer surplus as well as a the preconditions for effective competition—one of transfer of consumer surplus to producer surplus. If the main sources of gains—in the generation sector. the public sector were pricing below cost, the re- Each of the issues raised relates to system size. moval of this implicit subsidy would produce a “deadweight” gain, and a transfer back to taxpayers Costs and benefits of privatization and away from power consumers. Because subsidies Comparing the performance of a vertically integrated have been very large in many countries and the public monopoly with that of a vertically integrated state has financed new investment, moving from private monopoly exposes the gains and losses due public ownership to regulated private ownership to privatization alone. The key change that privatiza- (even if monopolistic) can produce large allocative tion introduces is the profit motive. The impact of net benefits for the economy. this change should not be confused with the effect of restructuring. In fact, the benefits to society as a This shift involves a potential gain in productive whole from ending state control of the power sector efficiency if private industry can cut costs. Public can be so large that the additional gains due to re- ownership tends to result in productive inefficiency, structuring may be relatively unimportant. both because managers have little incentive to re- duce costs and because politicians often are willing Performance assessments of publicly owned entities to increase costs to serve other purposes—for exam- should make a distinction between entities that have ple, providing secure employment. The political been corporatized and commercialized and those incentive to collect revenues or prevent theft of that have not been. Commercialization is possible power can also be low. only if the government removes itself from day-to-day interference in such issues as tariff setting and em- Whether a private monopoly will be productively ployment. Some countries that have not been ready to efficient (that is, produce a given output at mini- privatize their power sector have introduced commer- mum cost) is uncertain. The few well-established cialization (New Zealand, Portugal), an important inter- private monopolies (Barbados, Bermuda) appear to mediate step between the most interventionist form work well. The poor performance of many state com- of state ownership and privatization. Commercializa- panies is more likely to be attributable to the nature tion may allow many of the potential gains in effi- of their ownership than to their structure. ciency to be captured, especially where there is little scope for competition. Small systems may thus find it Costs and benefits of vertical separation of little incremental benefit to privatize, provided that Vertical separation—the separation of distribution, the government maintains an “arm’s length” relation- transmission, and generation into private monopolies— 2 has two important implications for private capital. It parts of the system. In a de-integrated structure, can increase monopoly power, but it can also lead the coordinating mechanisms are the prices and to the loss of economies of coordination. When contracts agreed between the two parties. Since each stage is monopolistic and the technology is each firm is trying to gain more of the profit for relatively fixed, a classic result is for an unregulated itself, there is a strong incentive in bargaining not chain of vertical monopolies to sell at a higher price to divulge information to the other, leading to than an unregulated integrated monopoly (the “dou- contracts that are suboptimal for the system as a ble wedge” problem). Regulation becomes more whole. In addition, there are transactions costs in important in this case. But the fact that many private negotiating and contracting. industries, even in competitive markets, show evi- dence of vertical integration indicates that there are A key argument in favor of separation is that it in- gains to be made from unified ownership, as in the creases transparency and allows the responsibilities U.S. power sector. of managers to become more focused. The larger the firm, the more difficult it is for a manager to The first general reason for the success of vertical have oversight of all its component parts and their integration is the existence of economies of scope. interrelationships. Certain activities need to be undertaken by both parts of the industry, but there is the possibility of Costs and benefits of horizontal separation sharing some inputs. A typical example would be in generation the need to have an accounting department in each The possibility of introducing competition into company. The activities of these departments would generation is critical to a power restructuring include handling the transactions between the two strategy. The key issue is the mechanism by which companies. Integration would not do away with competition takes place. If it is not possible to in- such transactions, although they become internal- troduce effective competition through vertical sep- ized, but they would be accounted for just once aration, on balance it may be better to leave the rather than twice. Related to this would be an econ- industry integrated even though it has been priva- omy of scale. An accounting department would not tized. Nor does the existence of several generating need to be twice as large to deal with a company companies by itself necessarily introduce competi- double the size. Some basic setup costs (computers) tion. So, breaking up the state company into sever- could be shared. A very important aspect of this ar- al private generators could lead to the loss of some gument is the existence of economies of scale to top benefits of scale or scope, yet without producing management. Good managers often are scarce (espe- any benefits through competitive downward pres- cially in small economies), and integration will likely sure on prices. save on this resource. There can also be financial economies of scale, which can be achieved when In large privatized power systems (Argentina, Chile, the component firms combine their borrowing England, and Wales), repeated bidding, in which the needs, reducing the cost of borrowing money. generators bid to supply power on a daily or even half-hourly basis, allows competition to be effective. A separate argument concerning vertical separa- If costs can be cut, then prices can immediately re- tion, or de-integration, relates to the decisionmak- flect this, forcing a higher-priced rival out of the ing process itself and economies of coordination. way. This bidding system is too complex for most This issue affects both day-to-day working of the smaller power systems and for economies at lower system (dispatch) and its longer-term size (invest- levels of development, which instead use a contract ment). In an integrated company, coordination system. With a long-run contract system, opportuni- takes place through physical commands and, to be ties for generators to use cost reductions to gain effective, requires complete information about all market share are much more infrequent. 3 A second condition for effective competition is that producers). If entry is easy and rapid, the threat of there be a sufficient number of firms to avoid implic- entry may be sufficient to induce existing firms to it collusion and gaming in the system. A two-genera- become cost-efficient. But where entry is difficult tor industry, for example, may be susceptible to (because of problems in obtaining licenses and con- each firm’s tacitly allowing its rival to behave in their structing the plants, for example), the threat of entry mutual interest. The arrangement in England and may be too small to affect the behavior of estab- Wales, with two large private generators (plus a lished firms. Where existing firms have some cost smaller, subsidized public nuclear company), has advantage not available to new entrants, there is an already demonstrated that a larger number of com- “intrinsic margin” that may not be competed away. panies is required to induce truly competitive behav- Common intrinsic advantages are privileged access ior. In small systems, the market can be too small to to local fuel supplies (especially hydro) that cannot support enough firms to achieve competitive condi- be bid away by higher contract prices for the fuel, tions—unless the firms are so small that they lose proximity to fuel source or to market, and environ- economies of scale. mental suitability of existing sites or even the non- availability of new sites. A third condition for effective competition is that the size and cost structure of the generating firms must In existing systems, de-integration may bring about be fairly similar. If they are not, there would be no some losses of economies of scale. One factor in possibility of using cost saving to increase market such losses is the need to maintain a “reserve mar- share (by altering the “merit order” with respect to a gin” against uncertainties. The experience of U.S. rival). A related issue in determining the competi- power pools has shown that pooling has enabled tiveness of rival generators is the “strength” of the individual firms to reduce reserve margins, and thus transmission system. A system with very high trans- to reduce costs. In developing countries, it is unlike- mission costs per unit of distance can allow some ly that such sophisticated devices can be made to generators to be virtual monopolies, since the extra work, so a de-integrated system will incur the extra costs of delivering supply across the transmission cost of maintaining reserve margins. Similarly, sepa- system to meet demand at a given node effectively ration will increase the demand for managers, who prevents competition from more “remote” sites. may be in scarce supply in many small and less de- veloped economies. In an existing industry, there must be excess capacity for competition to be successful in the short run. If A more detailed version of this Note will be produced in the Industry all plants are needed on a regular basis, there is no and Energy Department’s Occasional Papers series. incentive to cut costs. The force of competition must come from new entry (such as independent power Robert Bacon, Industry and Energy Department This series is published to share ideas and invite discussion. It covers financial and private sector development as well as industry and energy. The views expressed are those of the authors and are not intended to represent an official statement of Bank policy or strategy. Printed on recycled paper. Comments are welcome. Please call the FPD Note line to leave a message: 202-458-1111, or contact Suzanne Smith, editor, Room G8105, The World Bank, 1818 H Street, NW, Washington, DC 20433 or Internet address ssmith7@worldbank.org. 4